CPG-inspired workspace trajectory generation and adaptive locomotion control for quadruped robots

Area of Science:

• Robotics
• Biomimetic Engineering
• Control Systems

Background:

• Quadruped robots require sophisticated control for navigating complex and varied terrains.
• Existing control methods, such as traditional CPG-joint-space control, may lack adaptability to dynamic environmental changes.
• Biological systems offer insights into robust locomotion, particularly through central pattern generators (CPGs).

Purpose of the Study:

• To propose and validate a novel locomotion control architecture for quadruped robots inspired by biological CPGs.
• To develop an adaptive trajectory generation system capable of online parameter tuning for CPG networks.
• To demonstrate the superiority of the proposed control method over traditional approaches in diverse terrain conditions.

Main Methods:

• A control architecture comprising a 3-D workspace trajectory generator and a motion engine was designed.
• The workspace trajectory generator utilizes CPGs to produce adaptive trajectories by tuning network parameters online.
• A quadruped robot platform (AIBO) was employed for experimental validation of the proposed system.

Main Results:

• The proposed architecture successfully generated adaptive workspace trajectories, enabling online adaptation to various terrains.
• Experimental results confirmed the effectiveness of the control architecture in achieving robust locomotion.
• Comparative experiments demonstrated the proposed method's superiority over traditional CPG-joint-space control.

Conclusions:

• The developed control architecture provides an effective solution for adaptive locomotion in quadruped robots.
• Online tuning of CPG parameters allows for dynamic environmental adaptation, enhancing robot mobility.
• The proposed method represents a significant advancement in CPG-based robotic locomotion control.